Amazon integrated the open-source Ray framework into SageMaker HyperPod to streamline distributed Python workloads. This update combines Ray's scaling capabilities with HyperPod's automatic node recovery on Amazon EKS. It removes the manual overhead of managing KubeRay clusters. Practitioners can now deploy foundation model training and serving on resilient, purpose-built GPU infrastructure.